Neuroimaging research on emotional processing in major depressive disorder (MDD) yielded heterogeneous findings, characterized by small effect sizes when analyzing a priori defined emotional stimuli. In this study, we pursued a novel individual-level modeling approach to functional MRI data. N=275 participants — including healthy controls (HC, nHC=224) and patients with depression (nMDD=51) — completed an emotional processing paradigm. Blood-oxygen-level dependent responses were modeled for each individual using subjective emotional ratings as parametric regressors. The analysis revealed increased brain activation with increasing negativity rating in MDD compared to HC, with large to moderate effect sizes (d≤1·13) in widespread regions including the posterior cerebrum, basal ganglia, and limbic regions. Concluding, this innovative subject-level modeling approach extends previous research by demonstrating robust effects. The results not only show the strength of utilizing subjective evaluation compared to a priori defined emotional value, they also reveal neural underpinnings of altered emotional responsiveness in MDD.